Summary of an article published on The Web Scraping Club on 2023-03-13 (free to read). Read the full article

Kasada is one of the newest players in the anti-bot market, and it works differently from the systems most of us are used to. Wappalyzer does not identify it, so you have to recognize it from how the site behaves. The first request returns a 429 error instead of a CAPTCHA. If the client passes the check, the page reloads and you land on the original website. Kasada calls this Zero-Trust: every request is guilty until proven innocent.

I looked at what Kasada says about its own approach and at what the network tab shows when you browse a protected site. I put the two usual ways around an anti-bot next to each other: reverse engineering and deobfuscating the JS challenges, or making the scraper indistinguishable from a human browsing. Kasada’s claims about its proprietary interpretive language, polymorphic obfuscation and encrypted client communication bear directly on which way is practical. This is the first part. How to bypass it is covered in a follow-up.

What’s inside

  • How to recognize a Kasada-protected site when Wappalyzer does not detect it
  • Why the 429 on the first request is the challenge itself, with no CAPTCHA involved
  • Why reverse engineering the JS is good for learning but fragile against code that changes often
  • Which of Kasada’s own claims about detection on the first request matter for scraping

In the full article: Screenshots of the network tab on a Kasada site, the vendor quotes on Zero-Trust and polymorphic obfuscation, the embedded Kasada live demo video, and the link to the follow-up on bypassing it.

Read the full article on The Web Scraping Club